GENDER DIFFERENCES IN URBAN NEIGHBORING

Gender differences in neighboring are investigated using survey data from respondents in 81 Nashville neighborhoods. Of primary concern is why women neighbor more than men. Secondarily, the effects of life-cycle stage and socioeconomic status (SES) on neighboring are explored. Of these three, gender most strongly predicts neighboring activities, even with full and part time employment controlled. Life-cycle stage and SES are about equivalent in strength, although they significantly influence different neighboring measures. Findings suggest that the definition in American gender roles of women as bearers of emotional and social responsibilities accounts for their more extensive neighboring.     

A Methodology for More Efficient Tail Area Sampling with Discrete Probability Distributions

Monte Carlo Method is commonly used to observe the overall distribution and to determine the lower or upper bound value in statistical approach when direct analytical calculation is unavailable. However, this method would not be efficient if the tail area of a distribution is concerned. A new method, entitled Two-Step Tail Area Sampling, is developed, which uses the assumption of discrete probability distribution and samples only the tail area without distorting the overall distribution. This method uses a two-step sampling procedure. First, sampling at points separated by large intervals is done and second, sampling at points separated by small intervals is done with some check points determined at first-step sampling. Comparison with Monte Carlo Method shows that the results obtained from the new method converge to analytic value faster than Monte Carlo Method if the numbers of calculation of both methods are the same. This new method is applied to DNBR (Departure from Nucleate Boiling Ratio) prediction problem in design of the pressurized light water nuclear reactor.     

ON THE DERIVATION AND COMPUTATION OF THE CORNISH-FISHER EXPANSION

The paper gives a simple derivation based on elementary statistics for the Cornish—Fisher expansion. The method simplifies greatly the algebra in working out the polynomials involved both in the normalization and in the inverse expansions. The derivation also leads to recursive formulae which yield powerful algorithms for numerical computation of the expansion.     

False parsimony and its detection with GLMs

A search for a good parsimonious model is often required in data analysis. However, unfortunately we may end up with a falsely parsimonious model. Misspecification of the variance structure causes a loss of efficiency in regression estimation and this can lead to large standard-error estimates, producing possibly false parsimony. With generalized linear models (GLMs) we can keep the link function fixed while changing the variance function, thus allowing us to recognize false parsimony caused by such increased standard errors. With data transformation, any change of transformation automatically changes the scale for additivity, making false parsimony hard to recognize.     

Statistical morphology

This paper first introduces a parametric model for the generation of stationary random correlated binary sequences. The parameters of the model include the probability that a pixel is a binary one pixel and the length of the structuring element which dilates the initially spatially uncorrelated sequence. The spatial statistics of such eroded, dilated, opened and closed correlated binary sequences are derived in terms of the spatial statistics of the input binary sequence. Understanding of such one-dimensional processing is a precondition for understanding what happens in the more interesting two- dimensional case.